Epistasis occurs when the effect of one allele is influenced by another allele. Epistatic interactions play a profound role in evolution. An understanding of epistasis is also crucial in a variety of biomedical endeavors, such as identifying disease variants from genetic association studies and predicting the evolution of pathogens. But many of the most basic questions about epistasis remain unanswered, including: How common are epistatic interactions? How do they arise? And what are the underlying molecular mechanisms? We will address these questions by mapping epistatic interactions along a real evolutionary trajectory. To do this, we will employ a combination of computational and experimental tools to study the influenza nucleoprotein. Because of the unique nature of influenza evolution, we are able to infer in step-by-step detail the 39 mutations that have occurred in the nucleoprotein from human H3N2 since the year 1968. We will construct all 39 intermediate proteins along this evolutionary trajectory. We will also introduce each of the mutations individually into the 1968 parent. All of these variants will be tested for biochemical function and effect on viral fitness. This will provide a clear experimental test for epistasis: a mutation is involved in an epistatic interaction if it has a different effect in the 1968 parent than in the evolutionary intermediate in which it actually occurred. Crucially, our preliminary work has already identified several mutations involved in epistatic interactions. In Aim 1, we will build on this work by mapping all epistatic interactions since 1968. In Aim 2, we will address the mystery of how the epistatic interactions arose - did multiple mutations occur simultaneously, were there compensatory or permissive mutations, or did epistasis arise slowly due to a gradually shifting genetic background? Finally, in Aim 3, we will use biophysical and biochemical techniques to identify the molecular mechanisms of each epistatic interaction. At the conclusion of this study, we will have mapped the prevalence, origins, and mechanisms of epistatic interactions along a real evolutionary trajectory. Our findings will provide a new window into one of the most important factors shaping the evolution of proteins and viruses, and will aid in attempts to interpret sequence data and understand interactions between alleles.